Metal foreign matter detection system and detection method in wireless charging system

Abstract

The invention discloses a system and a method for detecting metal foreign matters in a wireless charging system, which are particularly suitable for detecting and positioning the metal foreign matters in the wireless charging system of an electric vehicle. The system comprises a transmitting coil, a receiving coil, a detection coil set, a conditioning circuit, a control unit and a BP neural network module. The detection coil set is of a two-dimensional double-layer structure and comprises a plurality of balance coil pairs, and each balance coil pair comprises two symmetrically-arranged detection coils and is used for detecting metal foreign matter in the high-frequency alternating magnetic field. And the conditioning circuit comprises a correction circuit, a filter circuit, an amplification circuit and a rectification circuit, and is used for processing the induced voltage signal output by the detection coil set and outputting a detection result. The control unit judges whether the metal foreign matter exists or not according to the output result of the conditioning circuit and determines the position of the metal foreign matter. The system has the advantages of simple structure, low power consumption, high integration level, low cost and the like.

Description

Technical Field

[0001] This invention relates to the field of wireless power transmission technology, and in particular to a metal foreign object detection system and method in a wireless charging system. Background Technology

[0002] With the rapid development of wireless power transfer technology, wireless charging systems for electric vehicles have gradually become a research hotspot. However, during the wireless charging process, the presence of metallic foreign objects may lead to safety hazards such as decreased system efficiency, excessive temperature rise, and even fire. Therefore, developing an efficient and accurate metallic foreign object detection system and method is of great significance.

[0003] Existing methods for detecting metallic foreign objects mainly include electromagnetic property detection, thermal sensing detection, radar and ultrasonic detection, and machine vision. Among these, electromagnetic property detection has become the preferred method for detecting metallic foreign objects in wireless power transmission systems due to its advantages such as simple structure, low power consumption, high integration, and low cost. However, existing electromagnetic property detection methods still have shortcomings in terms of detection accuracy, positioning capability, and system integration. Summary of the Invention

[0004] The purpose of this invention is to propose a metal foreign object detection system and method in a wireless charging system, so as to improve the positioning capability and detection accuracy of electromagnetic characteristic detection methods.

[0005] The technical solution to achieve the purpose of this invention is: a metal foreign object detection system in a wireless charging system, comprising:

[0006] The transmitting coil is used to generate a high-frequency alternating magnetic field;

[0007] A receiving coil is used to receive energy in a high-frequency alternating magnetic field;

[0008] The detection coil set includes multiple balanced coil pairs. Each balanced coil pair includes two symmetrically placed detection coils, namely the XY-axis detection coil and the ZW-axis detection coil, which are used to detect metallic foreign objects in a high-frequency alternating magnetic field.

[0009] The conditioning circuit amplifies and filters the induced voltage signal output from the detection coil, and outputs a distortion-free induced voltage signal.

[0010] The control unit is used to determine whether there is a metallic foreign object based on the induced voltage signal output by the conditioning circuit, and to determine the location of the metallic foreign object by the magnitude of the mutual inductance difference between the output XY axis detection coil and ZW axis detection coil.

[0011] The BP neural network module predicts the material type and size parameters of the metallic foreign object based on the mutual inductance difference between the XY-axis detection coil and the ZW-axis detection coil.

[0012] Furthermore, the transmitting coil, receiving coil, and detection coil set are all wound into a square shape using copper-coated Litz wire, with the receiving coil stacked on top of the transmitting coil and the detection coil set located between the transmitting coil and the receiving coil;

[0013] Furthermore, the detection coil set is a two-dimensional double-layer detection coil set, including: a first layer of detection coils, including multiple pairs of balanced coils arranged along the XY axis; and a second layer of detection coils, including multiple pairs of balanced coils arranged along the ZW axis, wherein the XY axis is orthogonal to the ZW axis.

[0014] Furthermore, each balanced coil pair in the detection coil set includes an XY-axis detection coil and a ZW-axis detection coil. The XY-axis detection coil and the ZW-axis detection coil have a non-overlapping symmetrical structure, and the voltage-free region width ratio of the XY-axis detection coil and the ZW-axis detection coil is 1.585.

[0015] Furthermore, the BP neural network module includes:

[0016] The input layer receives the mutual inductance difference between the XY-axis detection coil and the ZW-axis detection coil for metallic foreign objects.

[0017] The hidden layer uses the ReLU activation function for nonlinear transformation.

[0018] The output layer outputs the classification number and predicted size of the metallic foreign material.

[0019] Furthermore, the training process for the BP neural network module includes:

[0020] The mutual inductance difference between the XY-axis detection coils and the ZW-axis detection coils of metallic foreign objects of different materials and sizes was obtained through ANSYS simulation.

[0021] The weights and biases are updated using the backpropagation algorithm until the prediction error converges.

[0022] Through the coefficient of determination R 2 The root mean square error (RMSE) is used to evaluate the accuracy of a backpropagation (BP) neural network module, where the coefficient of determination R of the training set is used. 2 ≥0.97, test set R 2 ≥0.95.

[0023] A method for detecting metal foreign objects in a wireless charging system, based on the aforementioned metal foreign object detection system, enables the detection of metal foreign objects in the wireless charging system, comprising the following steps:

[0024] Step 1: Generate a high-frequency alternating magnetic field through the transmitting coil;

[0025] Step 2: Detect the induced voltage signal in the high-frequency alternating magnetic field using a detection coil set;

[0026] Step 3: The induced voltage signal output from the detection coil is amplified and filtered by the conditioning circuit, and a distortion-free induced voltage signal is output.

[0027] Step 4: The control unit determines whether there is a metallic foreign object based on the output induced voltage signal of the conditioning circuit, and determines the location of the metallic foreign object by measuring the mutual inductance difference between the output XY axis detection coil and ZW axis detection coil;

[0028] Step 5: The BP neural network module analyzes the mutual inductance difference between the XY axis detection coil and the ZW axis detection coil to predict the material type and size parameters of the metallic foreign object.

[0029] Compared with the prior art, the significant advantages of this invention are:

[0030] (1) By adopting a two-dimensional double-layer detection coil set, the energy transmission area of ​​the wireless charging system can be effectively covered, significantly improving the detection accuracy of metallic foreign objects. The balanced coil pair design of the detection coil set enables the system to sensitively capture minute changes in the magnetic field, thereby achieving accurate detection of small and medium-sized metallic foreign objects.

[0031] (2) The two-dimensional double-layer structure of the detection coil set, combined with the processing capability of the conditioning circuit, enables the rapid determination of the location of metallic foreign objects. By analyzing the mutual inductance difference between different detection coil pairs, the system can accurately locate the position of metallic foreign objects, reduce system downtime, and improve charging efficiency.

[0032] (3) The detection coil is wound with fine-diameter enameled wire, and the correction resistor in the conditioning circuit has a large resistance value, which significantly reduces the eddy current loss in the detection coil and reduces the overall power consumption of the system, meeting the requirements of energy saving and environmental protection.

[0033] (4) The compact design of the detection coil set and conditioning circuit makes it easy to integrate into existing wireless charging systems without significantly affecting the overall system structure. The high degree of integration makes its application in electric vehicle wireless charging systems more convenient.

[0034] (5) The filter circuit in the conditioning circuit can effectively filter out high-frequency noise and electromagnetic interference, ensuring the accuracy of the detection signal. The design of the amplifier circuit and rectifier circuit further enhances the anti-interference capability of the system, enabling the system to work stably in complex electromagnetic environments.

[0035] (6) The design of the detection coil set and conditioning circuit is simple, using common electronic components, which significantly reduces the manufacturing cost of the system. The low-cost design makes the system highly economical for large-scale applications.

[0036] (7) This system is not only applicable to wireless charging systems for electric vehicles, but can also be widely used in other wireless power transmission scenarios that require metal foreign object detection, such as medical equipment and industrial robots, and has broad application prospects.

[0037] (8) By detecting and locating metal foreign objects in real time, the system can issue alarms or take protective measures in a timely manner to avoid fires or other safety accidents caused by metal foreign objects, which significantly improves the safety of the wireless charging system. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the detection coil of the present invention;

[0039] Figure 2 This is a flowchart illustrating the detection principle of this invention;

[0040] Figure 3 These are the locations of sample points in the training set of the neural network of this invention;

[0041] Figure 4 These are the sample point locations in the neural network test set of this invention. Detailed Implementation

[0042] To make the technical methods, objectives, and functions of this invention easier to understand, the technical solutions involved in this invention will be clearly and completely described below in conjunction with specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this invention, and this invention is not limited to this specific embodiment.

[0043] A metal foreign object detection system in a wireless charging system, comprising:

[0044] The transmitting coil is used to generate a high-frequency alternating magnetic field;

[0045] A receiving coil is used to receive energy in a high-frequency alternating magnetic field;

[0046] The detection coil set includes multiple balanced coil pairs. Each balanced coil pair includes two symmetrically placed detection coils, namely the XY-axis detection coil and the ZW-axis detection coil, which are used to detect metallic foreign objects in a high-frequency alternating magnetic field.

[0047] The conditioning circuit amplifies and filters the induced voltage signal output from the detection coil, and outputs a distortion-free induced voltage signal.

[0048] The control unit is used to determine whether there is a metallic foreign object based on the induced voltage signal output by the conditioning circuit, and to determine the location of the metallic foreign object by the magnitude of the mutual inductance difference between the output XY axis detection coil and ZW axis detection coil.

[0049] The BP neural network module predicts the material type and size parameters of the metallic foreign object based on the mutual inductance difference between the XY-axis detection coil and the ZW-axis detection coil.

[0050] The transmitting coil, receiving coil, and detection coil set are all made of copper-coated Litz wire wound into a square shape. The receiving coil is stacked on top of the transmitting coil, and the detection coil set is located between the transmitting coil and the receiving coil.

[0051] Furthermore, the detection coil set is a two-dimensional double-layer detection coil set, including: a first layer of detection coils, including multiple pairs of balanced coils arranged along the XY axis; and a second layer of detection coils, including multiple pairs of balanced coils arranged along the ZW axis, wherein the XY axis is orthogonal to the ZW axis.

[0052] Furthermore, each balanced coil pair in the detection coil set includes an XY detection coil and a ZW detection coil. The XY detection coil and the ZW detection coil have a non-overlapping symmetrical structure, and the voltage-free region width ratio of the XY detection coil and the ZW detection coil is 1.585.

[0053] like Figure 1 The diagram shows a set of detection coils. Group X and Group Y are a pair of balanced coils in the vertical section of the diagram, while Group Z and Group W are a pair of balanced coils in the horizontal section. This two-dimensional detection coil set effectively improves detection efficiency. The area of ​​each pair of balanced coils should be equal, i.e., S... Xi =S Yi S Zi =S Wi (i = 1, 2, 3, 4, 5, 6).

[0054] The BP neural network module includes:

[0055] The input layer receives mutual inductance characteristic data of the metallic foreign object, including the mutual inductance difference between the XY axis detection coil and the ZW axis detection coil;

[0056] The hidden layer uses the ReLU activation function for nonlinear transformation.

[0057] The output layer outputs the classification number and predicted size of the metallic foreign object material.

[0058] The training process of the BP neural network module includes:

[0059] The mutual inductance difference between the XY-axis detection coils and the ZW-axis detection coils of metallic foreign objects of different materials and sizes was obtained through ANSYS simulation.

[0060] The weights and biases are updated using the backpropagation algorithm until the prediction error converges.

[0061] Through the coefficient of determination R2 The root mean square error (RMSE) is used to evaluate the accuracy of a BP neural network module, where the training set R... 2 ≥0.97, test set R 2 ≥0.95.

[0062] like Figure 2 As shown, a method for detecting metal foreign objects in a wireless charging system includes the following steps:

[0063] Step 1: Generate a high-frequency alternating magnetic field through the transmitting coil;

[0064] Step 2: Detect the induced voltage signal in the high-frequency alternating magnetic field using a detection coil set;

[0065] Step 3: The conditioning circuit amplifies and filters the induced voltage signal output from the detection coil, and outputs a distortion-free induced voltage signal.

[0066] Step 4: The control unit determines whether there are peak values ​​in the output induced voltage signals |Mxi|, |Myi|, |Mzi|, and |Mwi| based on the induced voltage signals output by the conditioning circuit. If there are peak values, there are metallic foreign objects; if there are no peak values, there are no metallic foreign objects.

[0067] Based on the mutual inductance difference between the output XY-axis detection coils and ZW-axis detection coils, the specific location of the metallic foreign object is determined by comparing the magnitude of the mutual inductance difference. If MAX|Mxi|>MAX|Myi|, the metallic foreign object is located in the X-axis detection coil; otherwise, it is located in the Y-axis detection coil. If MAX|Mzi|>MAX|Mwi|, the metallic foreign object is located in the Z-axis detection coil; otherwise, it is located in the W-axis detection coil, thus determining the specific location of the metallic foreign object. Step 5: The BP neural network module predicts the material type and size parameters of the metallic foreign object based on the mutual inductance difference between the XY-axis and ZW-axis detection coils.

[0068] Step 5 specifically includes:

[0069] The mutual inductance difference between the XY-axis and ZW-axis detection coil groups is used as the input feature;

[0070] The trained neural network model outputs the material classification probability distribution and size prediction values.

[0071] When the confidence level of material classification exceeds the preset threshold, an alarm for foreign object type is triggered.

[0072] Example

[0073] To verify the effectiveness of the present invention, the following simulation experiment was conducted.

[0074] like Figure 3 and Figure 4The images show sample point plots of the training and test sets for training a BP neural network in one embodiment. The training set contains 150 data points, and the test set contains 50 data points. For the 150 data points in the training set, the root mean square error (RMSE) is 0.011641. The predicted values ​​generally maintain good consistency with the true values ​​across the entire sample range, especially between 60 and 150 data points, where the best fitting effect is observed. For the first 50 data points, there is a slight deviation between the true and predicted values, with an error range within 0.1 nH, and the upper limit of the predicted values ​​is better than the lower limit. For the 50 data points in the test set, the RMSE is 0.014686.

[0075] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art should understand that within the scope of the technology disclosed in the present invention, any equivalent substitutions or changes made according to the technical solution and inventive concept of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A metal foreign object detection system in a wireless charging system, characterized in that, include: The transmitting coil is used to generate a high-frequency alternating magnetic field; A receiving coil is used to receive energy in a high-frequency alternating magnetic field; The detection coil set includes multiple balanced coil pairs. Each balanced coil pair includes two symmetrically placed detection coils, namely the XY-axis detection coil and the ZW-axis detection coil, which are used to detect metallic foreign objects in a high-frequency alternating magnetic field. The conditioning circuit amplifies and filters the induced voltage signal output from the detection coil, and outputs a distortion-free induced voltage signal. The control unit is used to determine whether there is a metallic foreign object based on the induced voltage signal output by the conditioning circuit, and to determine the location of the metallic foreign object by the magnitude of the mutual inductance difference between the output XY axis detection coil and ZW axis detection coil. The BP neural network module predicts the material type and size parameters of the metallic foreign object based on the mutual inductance difference between the XY-axis detection coil and the ZW-axis detection coil.

2. The metal foreign object detection system in the wireless charging system according to claim 1, characterized in that, The transmitting coil, receiving coil, and detection coil set are all made of copper-coated Litz wire wound into a square shape. The receiving coil is stacked on top of the transmitting coil, and the detection coil set is located between the transmitting coil and the receiving coil.

3. The metal foreign object detection system in a wireless charging system according to claim 1, characterized in that, The detection coil set is a two-dimensional double-layer detection coil set, including: a first layer of detection coils, including multiple pairs of balanced coils arranged along the XY axis; and a second layer of detection coils, including multiple pairs of balanced coils arranged along the ZW axis, wherein the XY axis is orthogonal to the ZW axis.

4. The metal foreign object detection system in a wireless charging system according to claim 3, characterized in that, Each balanced coil pair in the detection coil set includes an XY-axis detection coil and a ZW-axis detection coil. The XY-axis detection coil and the ZW-axis detection coil have a non-overlapping symmetrical structure, and the voltage-free region width ratio of the XY-axis detection coil and the ZW-axis detection coil is 1.

585.

5. The metal foreign object detection system in the wireless charging system according to claim 1, characterized in that, The BP neural network module includes: The input layer receives the mutual inductance difference between the XY-axis detection coil and the ZW-axis detection coil for metallic foreign objects. The hidden layer uses the ReLU activation function for nonlinear transformation. The output layer outputs the classification number and predicted size of the metallic foreign material.

6. The metal foreign object detection system in the wireless charging system according to claim 5, characterized in that, The training process for the BP neural network module includes: The mutual inductance difference between the XY-axis detection coils and the ZW-axis detection coils of metallic foreign objects of different materials and sizes was obtained through ANSYS simulation. The weights and biases are updated using the backpropagation algorithm until the prediction error converges. Through the coefficient of determination R 2 The root mean square error (RMSE) is used to evaluate the accuracy of a backpropagation (BP) neural network module, where the coefficient of determination R of the training set is used. 2 ≥0.97, test set R 2 ≥0.

95.

7. A method for detecting metallic foreign objects in a wireless charging system, characterized in that, The metal foreign object detection system according to any one of claims 1-6 enables metal foreign object detection in a wireless charging system, comprising the following steps: Step 1: Generate a high-frequency alternating magnetic field through the transmitting coil; Step 2: Detect the induced voltage signal in the high-frequency alternating magnetic field using a detection coil set; Step 3: The induced voltage signal output from the detection coil is amplified and filtered by the conditioning circuit, and a distortion-free induced voltage signal is output. Step 4: The control unit determines whether there is a metallic foreign object based on the output induced voltage signal of the conditioning circuit, and determines the location of the metallic foreign object by measuring the mutual inductance difference between the output XY axis detection coil and ZW axis detection coil; Step 5: The BP neural network module analyzes the mutual inductance difference between the XY axis detection coil and the ZW axis detection coil to predict the material type and size parameters of the metallic foreign object.

8. The method for detecting metal foreign objects in a wireless charging system according to claim 7, characterized in that, Step 4: The control unit determines the presence of a metallic foreign object based on the output induced voltage signal of the conditioning circuit. It determines the location of the metallic foreign object by measuring the mutual inductance difference between the XY-axis and ZW-axis detection coils. The specific method is as follows: Based on the induced voltage signal output by the conditioning circuit, determine whether the output induced voltage signals |Mxi|, |Myi|, |Mzi|, and |Mwi| have peak values. If peak values ​​are present, there is a metallic foreign object; if no peak values ​​are present, there is no metallic foreign object. Based on the mutual inductance difference between the output XY-axis detection coils and ZW-axis detection coils, the specific location of the metallic foreign object is determined by comparing the magnitude of the mutual inductance difference. If MAX|Mxi|>MAX|Myi|, the metallic foreign object is located in the X-axis detection coil; otherwise, it is located in the Y-axis detection coil. If MAX|Mzi|>MAX|Mwi|, the metallic foreign object is located in the Z-axis detection coil; otherwise, it is located in the W-axis detection coil. This method determines the specific location of the metallic foreign object.

Images